 |
Description  |
|
|
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention is directed to an automatic dilution system.
Specifically, the present invention provides for an automatic dilution
system which ideally provides for an exponential dilution to an
appropriate concentration.
2. Description of the Prior Art
The physical or chemical analysis of various types of fluid samples is
often accomplished by diluting the sample with a diluent to an appropriate
concentration. The fluid samples may be any of a wide variety of
solutions, suspensions, and dispersions. As used in the present
application, the term "diluent" refers to either a gas or liquid dependent
upon whether the dilution application requires the sample to be diluted in
a dry or liquid state.
In certain measurement applications, the dilution of the fluid sample is
carried out using a known dilution factor. This known dilution factor is
either predetermined at the onset of the measurement or is computed after
the dilution has been accomplished. In the prior art, a variety of methods
and devices have been developed to achieve a known or predetermined
dilution of a fluid sample. These prior art methods and devices have been
limited in their utility and cannot provide for a variable dilution when
such a variable dilution would be more useful than the known or
predetermined dilution of the prior art devices.
As an example, the following prior art patents may be pertinent to the
present invention. Cruzan U.S. Pat. No. 4,036,062, Roof et al U.S. Pat.
No. 4,036,063 and Roof U.S. Pat. No. 4,070,913 all describe means for
diluting a liquid sample with liquid diluent in which each of the two
fluids is initially contained in a pair of conduits. The two conduits are
connected together to permit a closed loop circulation and mixing of the
two fluids. The extent of the dilution is determined at the onset by
preselecting the volumetric relationship of the two conduits.
The Mowery, Jr. U.S. Pat. No. 4,095,472 describes a system wherein a liquid
sample is diluted by directing independent streams of a sample liquid and
a liquid diluent each at a constant preset flow rate into a mixing
chamber. The diluted sample fluid can then be extracted from the mixing
chamber. In this patent a fixed dilution factor is established at the
onset. The Culbertson U.S. Pat. No. 3,805,831 describes a mixing apparatus
for continuously and proportionally mixing one fluid stream with another.
The final sample concentration which emerges is determined by the
composition of each stream and their relative rates of flow.
The Pardikes U.S. Pat. No. 4,279,759 describes an optical sensing device to
measure the presence of a treatment chemical in a liquid process stream.
This patent also controls, by negative feedback, the rate of introduction
of the treatment chemical into the continuously flowing stream so as to
establish a relatively fixed concentration of the treatment chemical in
the stream. Moreaud et al U.S. Pat. No. 4,348,112, Tsuji et al U.S. Pat.
No. 4,408,880 and Brenholdt U.S. Pat. No. 4,507,556 describe various
sensor techniques based on light scattering and/or defraction to estimate
either the particle value or particle concentration in a liquid
suspension.
It can be seen, therefore, that a variety of methods and devices exist in
the prior art to achieve known or predetermined dilutions of a sample
fluid. However, there are other types of measurements of physical or
chemical properties of fluid samples wherein the measurement is more
properly accomplished by diluting the fluid sample to an extent which is
not predetermined at the outset of the dilution process. In these types of
measurements, the final extent of dilution may be controlled by some
measurable property of the fluid sample which property changes
considerably during the dilution process. For example, the measurable
property of the fluid sample may be optical turbidity, color, electrical
conductivity, pH, etc. The prior art methods and devices cannot provide
for this variable dilution which changes during the measurement process in
accordance with the change in some measurable property of the fluid
sample. The prior art methods and devices are limited in their utility for
this type of system.
There are a large number of commercial products which contain fine
particles which exist either in a dry state or suspended in an appropriate
solvent such as water. The physical and/or chemical properties of these
commercial products usually depend significantly on the distribution of
particle sizes or molecular weights of the individual particles or
molecules contained in the product. Typically, when liquid samples are
obtained in a manufacturing process, these samples contain a high
concentration of solute particles or macromolecules often exceeding 10%
concentration by weight or volume. However, most analytical instruments
are designed to measure particle size or molecular weight only if provided
with a sample in the form of a dispersion of particles in gas or liquid
which is much less concentrated than the concentration normally obtained
at the outset from the manufacturing process.
Therefore, there is usually the need to perform a substantial dilution of
the original sample. This dilution would normally be accomplished using a
fluid diluent which is either a gas or a liquid. For this type of
application and for others it would be desirable to develop a simple
dilution apparatus which yields an acceptable final dilution of a fluid
sample which is appropriate or optimal for the analytical measurement in
question. However, the dilution apparatus must ideally operate without any
knowledge of the starting concentration or composition of the particular
sample, whether in a dry state or in liquid suspension.
In the prior art, dilutions are normally achieved by measuring out a known
volume of a starting fluid sample into a suitable container and adding to
this, either simultaneously or subsequently, a known volume or amount of
diluent. The resulting mixture is then thoroughly mixed so as to disperse
the solute particles from the original sample uniformally within the new
fluid volume. The result ideally is a new fluid mixture or suspension
which is homogeneous and has a lower concentration of the solute component
then the original fluid sample at the onset of the dilution process.
As an example, U.S. Pat. Nos. 4,036,062, 4,036,063 and 4,070,913 describe
methods of carrying out such a fixed dilution. However, this traditional
approach to dilution is inconvenient and relatively inaccurate when large
dilution factors are desired. In these situations it is difficult to meter
out accurately a very small volume of starting sample fluid to be then
added to a given amount of diluent. To overcome this problem it may be
necessary to perform multiple dilutions in series in which each individual
dilution factor is relatively small and, therefore, accurately
controllable. The final dilution factor is then equal to the product of
the individual ones. However, such an apparatus is necessarily more
complex and more difficult to maintain because of the larger number of
individual stages.
In order to perform an analytical measurement a quantity of the new diluted
fluid sample is transferred from the mixing container into the appropriate
measuring instrument. This transfer is normally provided either by manual
means, such as pipetting, or by means of an automatic fluid handling
system. Unfortunately, for most analytical instruments the straightforward
method of diluting the fluid sample as described above is not very
efficient; rather, the dilution factor must often be adjusted in a
trial-and-error fashion in order to obtain a final dilution factor which
results in optimal performance of the analytical instrument. For example,
the initial dilution of the original fluid sample may be insufficient
thereby resulting in an overloading or saturation of the measuring
instrument. Alternatively, the dilution of the original fluid sample may
be too extensive thereby yielding an inaccurate measured signal.
Automatic dilution systems have also been developed which continuously
introduce both the starting sample and diluent fluid into the mixing
chamber. The input rates of each of these components can be adjusted to
fixed known values so as to yield a final diluted fluid sample whose
dilution factor remains known. The dilution factor may also be constant in
time as some of the final fluid sample is removed from the mixing valume.
These systems permit, at least in principle, the dilution factor to be
preset to any practical desired value to thereby result in a final solute
concentration ranging from a very low value to a very high value of
concentration. This type of adjustable dilution system may be seen with
reference to U.S. Pat. No. 4,095,472.
Automatic dilution systems have also been developed which rely on the
principle of negative feedback. In these systems, one or both of the flow
rates of the original sample and diluent into the mixing chamber are
continuously adjusted by a mechanism which responds to some measurement of
the resulting diluted fluid sample. Typical measurements include
turbidity, optical absorbance at a particular wavelength and light
scattering intensity, all of which are representative of the solute
concentration. The measurement which changes with the concentration of
solute particles in the diluent fluid sample is used to automatically
adjust the dilutor mechanism so as to yield an approximately unchanging
final solute concentration. Such a system is described in U.S. Pat. No.
4,279,759. This type of more sophisticated dilution system is actually an
adjustable version of the fixed dilution system described above. However,
because of the principle of negative feedback the final solute
concentration is kept approximately constant in time with the arrival of
addtional sample and diluent. The above described prior art automatic
dilution systems provide a background for the automatic dilution system of
the present invention which provides for an infinitly variable dilution of
a starting fluid sample.
SUMMARY OF THE INVENTION
The present invention is directed to an automatic dilution system and
provides for a method and apparatus to obtain a variable dilution of a
fluid sample with an appropriate diluent. This type of automatic dilution
is useful in applications in which a fixed predetermined dilution factor
is not required and, more importantly, is not useful. As an example, one
area of analytical measurement in which an automatic dilution system of
the present invention is appropriate is particle size analysis.
As an example, in the automatic dilution system of the present invention a
measured or unmeasured amount of concentrated original sample is
introduced into a mixing chamber. If the original sample is dry material,
the concentrated sample can be injected in the form of a quantity of dry
powder or a volume of gas in which is suspended dry sample at a high
concentration. If the sample is a liquid, then the input sample is a
concentrated liquid suspension of solute particles or molecules of the
sample. Diluent is then introduced into the same mixing chamber using
either the same input as used for the sample, or an additional input.
Typically for dry samples the diluent would consist of a pure gas or gas
mixture, such as air. For liquid samples, the diluent would be the same
liquid solvent used in the original concentrated sample, or perhaps a
different liquid.
As fresh diluent is introduced into the mixing chamber, mixed fluid exits
the chamber from an output which is preferably located at the furthest
point from the input of the starting sample and diluent. The exiting fluid
now contains a concentration of sample solute which is lower than the
concentration injected into the chamber at the input. This lowering of the
concentration is due to the dilution by the diluent flowing within the
mixing chamber. The sample solute concentration in the exiting fluid will
in general decrease with time as additional fresh diluent is introduced
into the chamber. Ideally the exiting concentration of solute decreases or
decays exponentially in time provided there is an ideal mixing of the
contents of the mixing chamber at all times and the rate of flow of fluid
through the chamber is constant.
The peak or maximum solute concentration which initially exits the mixing
chamber varies directly with the total amount of sample, S, which is
initially injected into the chamber and varies inversely with the volume,
V, of the chamber. Incomplete mixing results in a peak concentration value
which is either smaller or larger than the ideal value depending upon the
fluid flow characteristics within the mixing chamber. The characteristic
decay time, .tau., of the exponentially decaying solute concentration,
C(t), exiting the mixing chamber depends directly on the chamber volume,
V, and is inversely proportional to the rate of flow, F, of fresh diluent
into the chamber. All of these factors can be combined into a simple
mathematical expression for the solute concentration C(t) which exits the
mixing chamber as a function of time, t:
C(t)=S/V exp (-t/.tau.)
where
.tau.=
C(t)=S/V exp (-(F/V) t) Eq'n 1
The output fluid solute concentration C(t) can, for example, be expressed
in units of miligrams per cubic centimeter where the amount of injected
solute, S, is given in units of milligrams. The mixing chamber volume, V,
is then given in units of cubic centimeters and the rate of flow of fresh
diluent, F, is expressed in units of cubic centimeters per second.
The above equation 1 is valid for the idealized case in which the rate of
flow, F, of diluent into the mixing chamber is constant in time and the
contents of the chamber are thoroughly mixed at every instant of time.
This could also be expressed as the solute concentration being homogeneous
throughout the chamber. These conditions can be approximated if the mixing
chamber volume, V, is not vastly greater than the volume of sample
introduced and/or if the input diluent is injected with sufficient
velocity to induce turbulence in the chamber to result in a thorough
mixing of the chamber fluid contents by freshly arriving diluent.
Alternatively, some mechanical means of stirring the liquid contents of
the mixing chamber can be used to insure relative homogeneity throughout
the chamber.
Fortunately, the automatic dilution system of the present invention will
operate even if the dilution system does not behave in the ideal manner
described above. For example, if there is either a non-uniform
introduction of diluent or an incomplete or variable mixing of the chamber
contents, the output solute concentration, C(t), will not decay in time
following a simple exponential law for any interval. C(t) may decay
exponentially in time, but the characteristic decay time .tau. may vary
from one moment to the next depending on the input diluent flow rate and
the mixing characteristics of the mixing chamber thereby resulting in an
overall nonexponential rate of decay of solute concentration at the output
of the mixer.
In general, the solute concentration, C(t), exiting the mixing chamber will
decrease with increasing time as the finite amount of sample solute
originally introduced into the chamber is flushed from the chamber by the
continuous flow of fresh diluent. Whether or not C(t) falls monotonically
with time depends on the detailed mixing behavior of the chamber. However,
for a properly designed dilutor C(t) will in general decay approximately
exponentially with time.
The present invention yields, in principle, an infinitely variable dilution
of a starting fluid sample in which the dilution factor of the fluid
exiting the dilution device ideally decreases monotonically in time
following approximately an exponential decay law. The maximum
concentration obtained from this system is achieved at the beginning of
the dilution process, essentially at t=0. The more closely matched the
mixing volume is to the volume of starting sample initially injected into
the chamber, the larger is the initial solute concentration which exits
the chamber, C(t=0).
One limit for the system would be where the chamber volume, V, equals the
volume of injected sample fluid. With such a chamber volume, the output
solute concentration spans the maximum possible range, ranging from the
starting injected value to essentially zero after long elapsed times. If
on the other hand, the chamber volume, V, substantially exceeds the volume
of injected liquid sample, then the peak solute concentration which exits
the mixing chamber is substantially smaller than the starting volume. It
is, therefore, possible to choose the size of the mixing volume and the
volume of initially injected sample to obtain a substantial predilution of
an overly concentrated fluid sample, which predilution would occur before
the solute concentration is further reduced exponentially in time by the
dilutor.
Typically, a fluid sample which is injected into the mixing chamber
consists of a relatively small bolus, or pulse, of highly concentrated
solute/fluid suspension. As will be described in greater detail, the
dilution technique of the present invention can accommodate input sample
pulses of almost any volume or starting concentration. For example,
injections of the starting samples which are larger in volume and/or
higher in solute concentration than normal simply require a longer
dilution time to arrive at the same final solute concentration. With the
present invention, reproducable sample injections are not a prerequisite
for successful use.
In typical applications of the invention, the output from the mixing
chamber feeds a length of tubing or pipe which carries the diluted fluid
sample to a particular measuring or processing apparatus. Because the
exiting solute concentration, C(t), is in general decreasing monotonically
and exponentially in time, the solute concentration can be described as a
spatially varying function, C(x,t.sub.0), at a particular time (t.sub.0)
where x refers to linear distance along the tubing which carries the
output flow stream measured from the peak in solute concentration in the
direction of the static mixer. If the flushing diluent which enters the
mixing chamber is introduced at a constant flow rate and if there is
complete ideal mixing of the chamber contents at all times, as previously
described, then the spatial distribution of solute concentration at a
given time (t.sub.0), C(x,t.sub.0) is a monotonic decreasing function of
distance x. This function is an ideal exponential in distance x provided
there is not additional mixing of fluid within the tubing.
If it is desired to obtain a particular dilution of the original fluid
sample at some point (x.sub.0) or in some small region centered about
(x.sub.0) along the output flow stream, it is necessary only to stop the
flow of flushing diluent into the mixing chamber at the particular instant
of time, (t.sub.0), at which the solute concentration at (x.sub.0) has
decayed to the desired value. This solute concentration may be referred to
as C(x.sub.0, t.sub.0). At any point (x.sub.0), or small region centered
about (x.sub.0), any desired dilution of the original fluid sample is
obtained simply by waiting until the concentration C(x.sub.0, t) falls to
the desired value. The only requirement is that a sufficient quantity of
original sample, S, be injected into the mixing chamber so as to produce a
peak exiting concentration, equal to S/V, which exceeds the desired
concentration downstream from the mixing chamber output.
A very useful characteristic of the present invention is its potential for
relatively high speed and efficiency in achieving very large dilution
factors. Because the solute concentration which exits the mixing chamber,
or that at any point x.sub.0 along the output flow line decreases
exponentially in time, the desired solute dilution at any point (x.sub.0)
can be achieved relatively quickly even if an excessive amount of sample
is originally introduced into the mixing chamber. This desirable property
of the present invention will occur provided the system parameters of
mixing chamber volume, V, and diluent flow rate, F, are correctly chosen
to produce an exponential time constant .tau., equal to V/F which is
appropriately short, such as on the order of seconds.
For example, a dilutor system may be designed using values for V/F which
yield an exponential time constant of, as an example, three seconds. If it
is assumed that the sample solute concentration which initially exits the
mixing chamber is one hundred times larger than the desired concentration,
then after an elapsed time of three seconds after sample injection, the
exiting concentration will fall to 1/e or approximately 1/2.7 of the
initial exiting value. This is still a factor of 100/2.7, or 37, too
large. After two exponential time constants, equal to 6 seconds, the
exiting concentration is only 13.5 times too large. Finally, after only
4.6 .tau., or approximately 14 seconds, the exiting solute concentration
will fall to 1/100 of its original value which is the desired 100:1
additional dilution factor.
It can be seen, therefore, that the powerful nature of the exponential
function allows for a much larger initial concentration and still provides
for the desired dilution factor within a managable time period. For
example, even an initial overconcentration of a factor of 100,000 requires
a total elapsed time of only 11.5 decay constants to achieve the desired
final concentration. For the above example of .tau.=3 seconds, this
represents only about 35 seconds total time. Therefore, the automatic
exponential dilution system of the present invention can achieve fluid
sample dilution factors over an enormous dynamic range easily spanning a
range of 1,000,000 or greater.
These dilutions are achieved over a relatively short period of time without
the need to regulate accurately the flow rate of diluent over this time.
In the case of very large dilution factors, it may be useful to use a
variable flow rate system. For example, in the initial period following
sample injection, it may be appropriate to use a relatively high diluent
flow rate, F, which results in a small value for .tau. and rapid decay of
output solute concentration. When the solute concentration falls to
approximately the desired value, the diluent flow rate, F, may be
decreased in order to lengthen the time constant .tau. and thereby permit
more accurate monitoring of the solute concentration so as to be able to
halt diluent flow at the proper point in time.
BRIEF DESCRIPTION OF THE DRAWINGS
A clearer understanding of the present invention will be had with reference
to the following descriptions and drawings wherein.
FIG. 1 is a schematic drawing of a simplified embodiment of an exponential
dilutor;
FIG. 2 is a representative plot of output solute concentration exiting the
mixing chamber as a function of time;
FIG. 3 is a first embodiment of an exponential dilution system applied to a
light scattering based instrument;
FIG. 4 is a representative plot of the scattering intensity provided by the
embodiment of FIG. 3;
FIG. 5 is a second embodiment of an exponential dilution system appropriate
for on-line sample/particle sizing;
DESCRIPTION OF THE PREFERRED EMBODIMENTS
FIG. 1 is a schematic drawing of a simplified embodiment showing an
exponential dilutor. As shown in FIG. 1 a mixing chamber 10 may be defined
to have a volume equal to V. A sample injection may be supplied to the
mixing chamber 10 through an input port 12 and with the sample injection
provided through input tubing 14. The sample injection may be defined to
be an amount equal to S. A flow of diluent is provided through second
input tubing 16 which is also connected to the mixing chamber through the
input port 12. It is to be appreciated that in place of a single input
port 12, separate input ports may be used. The diluent input is provided
at a flow rate equal to F.
As fresh diluent continues to be introduced through the tubing 16 into the
mixing chamber 10, fluid exits the mixing chamber from an output port 18
preferably located at a point furthest away from the input port 12. The
exiting fluid may be passed through output tubing 20 and contains a
concentration of sample solute which is lower than the concentration
injected into the chamber through the input port 12 because of the mixing
with the diluent in the mixing chamber 10. The sample solute concentration
in the exiting fluid may be defined as being equal to C(t).
The exiting solute concentration as a function of time t is shown in FIG.
2. In general, the solute concentration of the exiting fluid decreases
with time as additional fresh diluent is introduced into the chamber 10.
Ideally, the exiting concentration of solute decreases exponentially in
time provided there is ideal mixing of the contents of the mixing chamber
at all times and the rate of flow of fluid through the chamber is
constant. As shown in FIG. 2, as sample and diluent are introduced into
the mixing chamber, the exiting concentration very rapidly builds to a
peak and then decays exponentially with time. The peak will vary directly
with the total amount of sample S, which is initially injected into the
chamber and will vary inversely with the volume of the chamber V. This is
shown in FIG. 2.
Before describing specific embodiments of the present invention, it will be
useful to consider the application of the exponential dilutor to a
particular measurement technology. This technology is particle size
determination using the method of Quasi-Elastic Light Scattering (QELS),
also known as Photon Correlation Spectroscopy. Diffusion coefficients of
particles suspended in liquid solution can be obtained from a mathematical
analysis of the autocorrelation function of the intensity of laser light
scattered at a particular angle which fluctuates in time due to the
Brownian motion of the particles. The particle diameter in turn is
obtained from the diffusivity by the Stokes-Einstein law. Typically,
colloidal suspensions which are analyzed by the QELS technique are much
too concentrated to be measured by a QELS-based instument and therefore
require substantial dilution, by a factor from 100 to 1,000,000 in most
cases. Dilution of the starting liquid sample is required to reduce the
level of light scattering from the suspended solute particles and thereby
avoid serious degradation of the performance of the particle sizing
instrument due to a variety of physical/optical effects.
Therefore, a dilution system which is appropriate for a QELS-based
instrument is one which produces a liquid sample which yields on average a
particular level of scattered light from a laser source of fixed
intensity. The exponential dilutor system described in the present
application is suited to this type of measurement. Specifically, the
diluted liquid sample which exits the mixing chamber is directed into a
flow-through scattering cell within the instrument. The fluid which exits
the instrument cell is discarded or recycled if desired to recover the
solute component. The proper dilution factor is obtained by continuously
monitoring the light scattering intensity obtained from the measurement
cell as the diluted liquid sample flows through the cell. This intensity
varies with the solute concentration C(x.sub.0, t) described above, where
position variable x.sub.0 refers to the measuring point along the output
flow stream. When the light scattering intensity falls to the desired
level appropriate for an autocorrelation measurement of the particle size,
the flow of liquid diluent into the mixing chamber is halted. A particle
size measurement may then be accomplished by analyzing the fluctuations in
the scattered light intensity which originate from the diluted fluid
sample within the same measurement cell. After a measurement has been
completed, additional diluent may be flowed through the system to flush
out any remaining sample solute from the mixing chamber and output
line/scattering cell.
A specific application of the exponential dilutor system of FIG. 1 to a
light scattering based instrument, such as QELS, is shown in FIG. 3.
Components of the system of FIG. 3 which are substantially the same as
that shown in FIG. 1 are given the same reference characters. Concen | | |
